Two days ago we got alerts about fluent-bit got oom killed.
While checking logs everywhere trying to find the reason I found the error log message in opensearch during the issue time frame on two of three nodes:
Lock is null. Nothing to release.
This message repeated 313 times during 30 minutes.
I can’t find any other issues in the logs that might explain what’s happening so this message is telling me nothing.
What does this message actually telling me about our opensearch cluster?
Thank you @cwperks for that information.
Why does that error occur? We’ve seen the error occasionally that’s why my primary question. I suspect we have a network issue and that’s probably the reason we got the oom kill of fluent-bit.